Abstract: A genetic-particle swarm optimization hybrid algorithm (GA-PSO) is proposed to address the low solution efficiency and poor dynamic adaptability of traditional methods in optimizing complex coal mine ventilation networks. An optimization model is established,constrained by the basic laws of mine ventilation and fan characteristic curves,with the objective of minimizing ventilation power consumption. To overcome the slow convergence rate of GA,a combination of stochastic competition and arithmetic cross-Gaussian mutation operators was employed to improve population diversity,enhance global convergence and avoid local optimization. In view of the precocious phenomenon of PSO, an elimination strategy integrating potential particle replacement and redundant particle restart mechanisms was designed,and the adaptive inertial weight adjustment strategy based on the standard deviation of adaptive values is proposed to improve the global search ability of the algorithm. Aa dynamic coordination mechanism of learning factors was combined to achieve a dynamic balance between global exploration and local optimization. Results demonstrate that the GA -PSO algorithm reduces ventilator power consumption by 16. 86% and restricts airflow errors within 4. 85%,outperforming standalone GA or PSO in convergence speed and optimization accuracy. This method effectively resolves challenges of premature convergence and high -dimensionality complexities in traditional ventilation network analyses, offering a theoretical foundation for energy-efficient and safe regulation of mine ventilation systems.